array9

The main issue would be finding a reasonably quiet location.  You would place your mobiles in some simple parabolic reflector pointing roughly to the same area of sky.  This probably does not have to be very exact, as it main purpose is to cut out local noise.

Then just record sounds on your mobiles, after you have synced their clocks.  This does not have to be hugely accurate, as you are only wanting to locate noises roughly in three dimensions.

You are looking for bird calls in a volume of space.  So if all devises pick up a similar call within a second or two of each other, you can calculate its position in space from the recorded timings.  Confirmation can come from comparing the sounds length, volume and shape.  

Most of this can be done totally automatically.  You can then manually investigate the interesting events, and see if they can be identified.  If lucky you may have a number of records for the same object and see the speed and direction of flight.

This not much more complicated than a group of people hearing a bird call in flight, and then discussing where you thought it was.   I think the amount of processing is quite within mobile phones ability to achieve in real time.

The info is not complete, in that you dont record silent birds, but would be interesting.

Food for thought. A student group project?

Martin